Abstract Recovery involves replacing drinking with more adaptive behavior. Once the more adaptive behavior becomes habitual, recovery is more resistant to relapse produced by stimuli previously occasioning drinking. However, until this transition occurs, the new adaptive behavior is easily displaced. This transition requires cognitive flexibility in order to attend to the stimuli and contingencies associated with the new behavior. Unfortunately, heavy drinking reduces cognitive flexibility, especially in females. This reduced cognitive flexibility appears to result from reduced cholinergic activity at ?4?2 and ?7 nicotinic receptors, particularly those in the hippocampus and anterior cingulate innervated by cholinergic neurons originating in the medial septum. Restoring cholinergic function may restore cognitive flexibility during the new learning involved in transitioning from habitual drinking to more adaptive habits; and thus, speed this transition. Speeding this transition minimizes the most fragile period of recovery, the early stages before new habits have developed. However, increasing cholinergic function above an optimal level impairs new learning. Thus, enhancing cholinergic function may only speed recovery when chronic drinking has reduced cognitive flexibility, perhaps more so in females. We explore these possibilities using a rat model of recovery we developed. Like in the real world, in this model high and low levels of drinking are controlled not by changing ethanol availability, but rather by changing the contingencies governing the availability of alternatives, in this case food. In the presence of one stimulus, food is costly and drinking predominates. In the presence of another, food is easily available and little drinking occurs, simulating recovery. After several consecutive recovery sessions in which rats only encounter low cost food conditions, presenting the high-cost food stimulus no longer controls responding for ethanol. Instead rats persist at responding for food, indicating this response has become habitual and control by the high-cost food stimulus has weakened. We propose examining whether chronic intermittent ethanol exposure slows the development of recovery, and whether this effect is greater in females. Further, we compare effects of nicotine to those of epibatidine and PNU-282987, which preferentially activate ?4?2 and ?7 receptors, respectively. Finally, we examine effects of delivering nicotine into the hippocampus and cingulate and compare these to those following administration into the thalamus (not expected to influence cognitive flexibility). This project explores an innovative target for relapse prevention: medications intended to speed new habit formation, such as might occur during behavioral therapy. This strategy may reduce relapse during the early, critical period of recovery, thus improving therapeutic outcomes.